At the present time, the unpublished line list for both Cf I and Cf II totalizes about 13000 entries in the whole spectral range. In 1989, the analysis of Cf I led to 129 even levels and 233 odd levels. It comprises in the even parity, 13 levels of the ground configuration and some of the lowest levels of 5f106d7s7p, 5f96d7s7p, 5f107s8s, 5f107s7d and probably 5f107s9s. The odd configurations overlap and mix, the lowest level of 5f107s7p being located only 550 cm-1 above the first excited odd level 5f9(H 15/2)6d3/27s2 J = 8. The relative hyperfine structure width of the levels (the width of the ground level being assumed to be zero) has been determined in most of the cases, as well as the g-factor. The combination of a small nuclear magnetic moment (-0.281 ± 0.06 nm) (86.8) and of a probably large electric quadrupole moment made it difficult the interpretation of many hyperfine structures. The three-place level values given to the nearest 0.005 cm-1 are accurate to ± 0.010 cm-1, while the two-place level values and the widths are accurate to ± 0.05 cm-1 or better. The measurment of a few isotope shifts 249Cf - 250Cf on the very strong lines was used also for identifying the energy levels empirically, as well as observed intensities and regular trends of electronic jumps along the 5f-period. This step in the analysis was reported in the Tables of Actinides (1992) as (89.1).The later publication (95.1) reported seven even parity levels and 32 odd parity levels which supplement the reference (89.1).
The theoretical investigations published for the Cf I levels are limited so far to the 5fl07s2 configuration (80.4). Complete studies of 5f107s7p and 5f96d7s2 led to a few additional identifications (unp.32).